What are the differences among centWave, MatchedFilter and Massifquant (peak picking)?

MetaboAnalyst currently supports three algorithms for LC-MS peak picking. They are used for different scenarios, as illustrated in details below,

  • centWave, an algorithm which was created by Ralf T. et al, 2008 has been extensively used for peak picking on high-resolution mass spectrometry data. It is using a continuous wavelet transformation model and optionally Gauss-fitting in the chromatographic domain. This algorithm is being used by default, and usually shows statisfying performance.

  • Matched Filter is mainly designed for low-resolution MS. This algorithm is being depreciated.

  • Massifquant is an Kalman filter-based algorithm, developed by Christopher J. et al, in 2014. This algorithm is quite sensitive towards the ms signals, and should be used for MS data with low intensity or the MS data was acquired with too many MS2 scans simultaneously.